Cyril Mango

A FAKE INSCRIPTION OF THE EMPRESS EUDOCIA
AND PULCHERIA’S RELIC OF SAINT STEPHEN
In her thorough discussion of St. Peter’s Chains Vera von Falkenhausen
has dwelt on the legend associating that relic with the Empress
Eudocia, wife of Theodosius II 1 . In fact – if one can speak of facts in
such matters – the learned Empress had nothing to do with the famous
Chains, but she did bring from Jerusalem a relic of St. Stephen, whose
story I shall attempt to disentangle in the following note.
In or shortly before 1889 the French scholar G. Doublet visited the
town of Safranbolu in Paphlagonia and was shown in the Greek church
of St. Stephen the copy of an inscription. «La pierre», he says, «a disparu».
He published the copy as follows 2 :
Σωτρ φανες, Στέφανε, λγεινν πνων
λαι γνατς κα πδς κτρς φίλης,
θεν ναν δωρµαι 3 κλεινι τι πλει
τ Θεδώρυ, κράντρς παλαιφάτυ,
5 δωρυµένη ληφθέντα δρν σν πδα
ατι µένειν, σσσηµν λήστυ µνείας
ΒΣ Εϒ∆ ΦΙΕ ΘΑΡΓ
Here is Doublet’s translation: «O Saint-Etienne! toi qui visiblement
m’as sauvée des cruelles douleurs que je souffrais, moi, ta misérable amie,
dans le genou gauche et dans le pied, je fais don de ce temple divin à la
glorieuse ville de Théodore, l’illustre guerrier. Et ton pied que j’avais
reçu moi-même en cadeau, je le donne à cette église, afin qu’il y reste et
que l’on s’en souvienne à jamais».
For the interpretation of this text Doublet turned to L. Duchesne,
1 V. VON FALKENHAUSEN, Petri Kettenfeier in Byzanz, in Fest und Alltag in Byzanz,
ed. G. PRINZING - D. SIMON, Munich 1990, pp.129-144.
2 Inscriptions de Paphlagonie, in Bulletin de Correspondance Hellénique 13 (1889), p.
294f. 3 As we shall see, the «correct» reading is δµµαι.
23

24
CYRIL MANGO
who declared the last line suspect, especially the figure (date?) 515 and
the Attic month Thargelion, but was satisfied with the rest of the inscription,
which he ascribed to Eudocia, wife of Theodosius II. He also concluded
that Safranbolu, being the city of Theodore, was the same as
Euchaita, and if the latter was renamed Theodoroupolis only in the reign
of John I Tzimiskes 4 , it may have been known earlier by that name on
an unofficial basis.
Once launched into print with Duchesne’s blessing, the Safranbolu
inscription has enjoyed and continues to enjoy remarkable success. William
Ramsay was greatly exercised by it and devoted several pages to
demonstrating that Safranbolu could not have been Euchaita. The latter,
he opined, was at Çorum (not quite right, but closer to the truth) 5 . The
inscription was accepted by H. Leclercq 6 , H. Delehaye, who, however,
explained that it did not refer to Euchaita 7 , F. Halkin 8 , D. Gorce in his
edition of the Life of St. Melania 9 , E.D. Hunt 10 , K.G. Holum 11 , P. Maraval
12 and, with complete assurance, by E. Livrea 13 . It is also, presumably,
for the same reason that on the Classical Map of Asia Minor by W.M.
CALDER and G.E. BEAN (London 1958) the site of Safranbolu is marked
Germia Theodorias. Scholars who have expressed scepticism have been
fewer: I can name only the obscure Gabrielides (see below), Chr. Marek
14 , K. Belke 15 , R. Merkelbach and J. Stauber 16 . D. Feissel has prudently
4 Actually, the name Theodoroupolis appears to have been given to Euchaina or
Euchaneia (Çorum), often confused with Euchaita. See N. OIKONOMIDES, Le dédoublement
de S.Théodore et les villes d’Euchaita et d’Euchaneia, in Analecta Bollandiana 104
(1986), pp. 327-335.
5 The Historical Geography of Asia Minor, London 1890, pp.21, 52-53, 319-322.
6 Art. Eukhaita, in Dictionnaire d’Archéologie Chrétienne et de Liturgie, V/2, Paris
1922, pp.704-705.
7 Acta Sanctorum, Nov. IV (1925), p. 23.
8 Inscriptions grecques relatives à l’hagiographie, in Analecta Bollandiana 71 (1953), p. 96.
9 Vie de Sainte Mélanie, Paris 1962 (Sources Chrétiennes, 90), p. 246 n. 1.
10 Holy Land Pilgrimage in the Later Roman Empire, AD 312-460, Oxford 1982,
p. 233.
11 Theodosian Empresses, Berkeley 1982, p.186 n. 52.
12 Lieux saints et pèlerinages d’Orient, Paris 1985, p.368.
13 La slogatura di Eudocia in un’iscrizione paflagone, in Zeitschrift für Papyrologie und
Epigraphik 113 (1996), pp. 71-76.
14 Stadt, Ära und Territorium in Pontus-Bithynia, Tübingen 1993 (Istanbuler Forschungen,
39), p. 191 nr. 11: «Die Abschrift scheint mir verdachtig».
15 Paphlagonien und Honorias,Vienna 1996 (Tabula Imperii Byzantini, 9), p. 268:
Probably of the 19th century, but does it refer to the wife of Theodosius II or to
Eudocia Makrembolitissa, wife of Constantine X and Romanus IV?
16 Steinepigramme aus dem griechischen Osten, II, Munich 2001, p.308 nr. 10/
02/99: Metrical faults too serious.

A FAKE INSCRIPTION OF THE EMPRESS EUDOCIA 25
suspended judgment 17 .
None of the above excellent scholars appears to have asked the obvious
question: What was Eudocia doing at Safranbolu? Did she go there
on holiday? Assuming she was in possession of an important relic of St.
Stephen (in addition to the one, as we shall see, that she was to give to
her sister-in-law Pulcheria), she must have been passing through those
parts on her return journey from Jerusalem in 439. It is reasonable to
assume that she would have followed the so-called «Pilgrim’s Road»,
which cut diagonally across Asia Minor in the direction of Antioch.That
road, however, was nowhere near Safranbolu 18 . We must assume, therefore,
that on a whim she took a lengthy detour to visit an obscure locality
(ancient name unknown) and, once there, built a church in honour
of St. Stephen to which she donated her precious relic.
Safranbolu, which has few, if any ancient remains, has not attracted
much attention on the part of European travellers. A.D. Mordtmann,
who visited it in 1856, says that it passed for («gilt für») medieval
Theodoroupolis. It had a separate Greek «ghetto» called Granköi (read
Kıranköy) of about 250 families with a church of St. Stephen, built in
1805. He was shown there a relic of St. Stephen’s right (sic) foot, on
whose authenticity he was unwilling to pass judgment. No word about
an inscription 19 . The Greek settlement multiplied – it is said to have
numbered 3,000 before the exchange of populations in 1922 – and in
1871 built a much bigger church of St. Stephen with a dome and two
bell-towers. It must have been the new church that Doublet visited. He
failed, however, to see in it another inscription – a genuine epitaph, that
was probably brought in from the surrounding countryside 20 . Eventually
St. Stephen’s was turned into a mosque, which is still standing minus the
bell-towers, and on either side of the central door is our Eudocia
inscription, carved twice in neat 19th-century characters, in capital letters
on the left and lower case on the right (Pl. 1-2) 21 . It differs from
17 Bulletin épigraphique, 1998, No.659 = Revue des études grecques 111 (1998),
p. 711.
18 See D. FRENCH, Roman Roads and Milestones of Asia Minor, Fasc. 1: The Pilgrim’s
Road, Oxford 1981 (BAR Internat. Series, 105), map 4 (Zonguldak sheet).
19 A.D. MORDTMANN d. Ä., Anatolien. Skizzen und Reisebriefe aus Kleinasien
(1850-1859), ed. F. BABINGER, Hannover 1925, p.253f.
20 Published by E. LEGRAND, Inscriptions de Paphlagonie, in Bulletin de Correspondance
Hellénique 21 (1897), p. 92.
21 For the photographs reproduced here, taken by Ms. Vildan Gök, I am very
grateful to Dr. M.I. Tunay. Poor photo in Chr. KYRIAKOPOULOS, Σαφράνπλη -

A FAKE INSCRIPTION OF THE EMPRESS EUDOCIA 27
Doublet’s copy in having δµµαι instead of δωρµαι in line 3 and
ΘΑΡ.Η instead of ΘΑΡΓ in the last line, conveniently glossed into Karamanlidika:
πς γι νπς σενεσιντέ, πριλτ η («In the year 515, April 8»).
I do not know why Thargelion (May/June) was thought to be April.
One does not have to be an expert in Late Antique poetry to see
that the Eudocia inscription belongs, not to the world of the 5th century
22 , but to that of 19th-century ellinismos. It is in «Byzantine» dodecasyllables
with little regard for the quantity of vowels and starts with a
nominative absolute (φανείς), i.e. Stephen, followed by the first-person
δµµαι, i.e. Eudocia.These incongruities were pointed out long ago by
one A. Gabrielides, who ascribed the poem to a «half-educated schoolmaster»
(µιµαθς διδάσκαλς) 23 . Attention should also be drawn to the
rare word κράντωρ (line 4), which Doublet was mistaken to translate
«guerrier» and Livrea as «eroe», «santo». Κράντωρ means «ruler», princeps
or effector (as in the epigram quoted by Pausanias, 8.52.3), and is correctly
defined in what was probably the standard Greek dictionary for Greeks
that the «half-educated schoolmaster» might have consulted 24 . In other
words, he described Theodore as a ruler, not as a warrior saint. The
remarks of the worthy Gabrielides may throw some light on this little
puzzle. After saying that in his day Safranbolu was called Theodoroupolis,
he adds that he did not know who the Theodore in question was. An
old schoolmaster (a different one?) told him the following story. The
Empress Eudocia, wife of Theodosius, fell ill and proceeded to that town
for a change of air in the year 515 (sic). There she got better and attributed
her cure to the left foot of St. Stephen, which she had brought with
her. She made a gift of the relic to the town, whose governor was called
Theodore. Gabrielides adds that the relic was still preserved and extended
from the left knee to the tip of the foot – quite a sizeable assemblage
of bones. In other words, the inscription reflected the local legend
as it was told in the 19th century and the nature of the relic that was
kept there at the time.
∆άδυρα (sic), 2nd ed., Thessaloniki 1988, fig. 6 on p. 201. For the Eudocia inscription,
ibid., pp. 88, 101 and fig. 7 on p. 202. I am indebted to Dr.Vasso Pennas for a
copy of this somewhat incoherent publication.
22 The contrast with the genuine hexameter inscription of Eudocia in the baths
of Gadara (Suppl. epigr. gr. 32, No.1502) could hardly be more striking.
23 Περ τς παρίας Νεκαισαρείας, in ενφάνης 1 (1896), pp. 132-133. I am indebted
to D. Feissel for bringing this scarce publication to my attention.
24 Λεικν τς λληνικς γλώσσης πίτµν, ed. K. GARPOLAS, Athens 1839.

28
CYRIL MANGO
There remains a curious coincidence that has not gone unnoticed.
Eudocia did dislocate her foot during her first visit to Jerusalem (438-
439) and was allegedly healed by the prayers of St. Melania offered in
front of certain unspecified relics.The story is told in the original Life of
Melania 25 , whose Greek text was first published in 1903, and so could
not have been known to the author of the Safranbolu inscription. He
might, however, have been aware of the Metaphrastic version, which
repeats the same tale 26 . No source known to me specifies that Eudocia
injured her left knee and foot.
After eliminating from the dossier the inscription of Safranbolu, we
may consider a related incident that has to do with St. Stephen’s relics,
this time at Constantinople. First, however, we may set down certain
facts that appear to be reasonably well established:
December 415: St. Stephen’s relics are found at Kapher-Gamala near
Jerusalem and are temporarily deposited in the Sion church. News of the
discovery echoes throughout the Christian world and some dispersion of
small fragments, notably to the West, immediately takes place 27 .
c. 416: Aurelian, Praetorian Prefect of the East, 399, 414-416 (i.e. at the
very time of the invention) builds a church in honour of the Saint on
his estate at Constantinople in the expectation of obtaining some of the
relics, but his wish is not satisfied, which is rather odd, considering his
influential position 28 .
439: At Jerusalem a martyrium of St. Stephen outside the north gate
of the city, built either by Eudocia or bishop Juvenal (or both) is dedicated
on 15 May. The Saint’s remains (presumably the major part of
them) are deposited there 29 . On returning to Constantinople the same
year, Eudocia gives her sister-in-law Pulcheria a relic of St. Stephen
(body part unspecified), which is laid in Pulcheria’s basilica of St. Laurence
at Blachernai 30 .
25 Ed. GORCE, cit., p. 244.
26 PG 116, col. 788: Τν γρ πδα τς ασιλίδς πηρεία δαίµνς τν ρµνίαν
παρατραπέντα κείνη [Melania]… ε µνη πρς τ εθς πανήγαγε.
27 See S. VANDERLINDEN, Revelatio sancti Stephani, in Revue des études byzantines 4
(1946), pp. 178-217.
28 See A. CAMERON - J. LONG, Barbarians and Politics at the Court of Arcadius,
Berkeley 1993, p.74. The source (Vita Isaacii) is relatively late, but appears credible.
29 See, e.g., E. HONIGMANN, Juvenal of Jerusalem, in Dumbarton Oaks Papers 5
(1950), pp. 225-227.
30 MARCELLINUS COMES, Chronicon, s.a. B. Croke’s comment on this passage in
The Chronicle of Marcellinus, Sydney 1995, p.83 is somewhat confused.

A FAKE INSCRIPTION OF THE EMPRESS EUDOCIA 29
Mid-9th century or earlier 31 :The cult centre of St. Stephen at Constantinople,
including his relics, has shifted to the district of Constanti(ni)anae.
The problem I should like to examine is whether there occurred a
translation to Constantinople of a major relic of St. Stephen, namely his
right arm or hand, as early as 427/428 (or possibly 421). The only authority
for such a translation is the Chronicle of Theophanes, which
under AM 5920 (= AD 427/428) relates the following story. In return for
a large sum of money and a gold jewelled cross to be erected on the site
of Calvary, the archbishop of Jerusalem (unnamed, but given as Praylius
in the rubric) sent to Constantinople Stephen’s right είρ (let us call it
arm).The relic was conveyed to Constantinople by the holy man Passarion
32 . Pulcheria, apprised in a vision, went out to meet it and built a
wonderful church (κν νδν) in the Imperial Palace wherein she
deposited it 33 .
The date 427/428 is certainly not to be taken very seriously. Theophanes’
chronology for the relevant period is hopelessly muddled: he
places the invention at Kapher-Gamala in 426/427 instead of the correct
415 and is completely mixed up about the succession of Jerusalem’s bishops.
Having no other material for AM 5920, he may simply have inserted
there the story of the relic’s adventus, which he derived, perhaps, from a
hagiographic source. A. Frolow arbitrarily changed the date to 420 34 ,
whereas K. Holum, for reasons of his own, has argued vigorously that the
incident occurred, not in 427/428 (the 20th year of Theodosius II counting
from his accession in 408), but in 421 (the 20th year from his proclamation
as Augustus on 10 January 402) 35 .
There can be no doubt that a chapel of St. Stephen did exist in the
Imperial Palace, more precisely in the part of the palace called Daphne.
It is attested in the reign of Zeno (474-491) when it received a copy of
31 Seeing that the supremely silly translatio to Constanti(ni)anae was rendered
into Latin by Anastasius Bibliothecarius. Text in PL 41, col. 817f.
32 Passarion died in extreme old age in Oct./Nov. 428: CYRIL OF SCYTHOPOLIS,
Vita Euthymii, ed. E. SCHWARTZ, Leipzig 1939 (Texte und Untersuchungen zur
Geschichte der altchristlichen Literatur, 49/2), pp. 26-27.
33 Ed. C. DE BOOR, pp.86-87; repeated by CEDRENUS, Bonn ed., I, p. 592.
34 Numismatique byzantine et archéologie des Lieux Saints, in Mémorial Louis Petit,
Bucarest 1948, p.84.
35 Pulcheria’s Crusade, AD 421-22, in Greek, Roman and Byzantine Studies 18
(1977), p. 163. See further K.G. HOLUM - G. VIKAN, The Trier Ivory, in Dumbarton
Oaks Papers 33 (1979), esp. p. 127f.

30
CYRIL MANGO
St. Matthew’s Gospel written in the hand of Barnabas that had been
found in Cyprus 36 , an incident dated by Victor Tunnunensis to 488 37 .
From the reign of Heraclius onwards the chapel in question is repeatedly
mentioned as a venue for imperial coronations and weddings 38 and it
figures prominently in the Book of Ceremonies 39 . There is, however, no
independent confirmation that it had been built by Pulcheria. That, it
may be thought, is not too serious an objection, but what is odd is that
the famous dexia plays no part whatever in imperial ceremonial 40 , indeed
is never alluded to, unless I am mistaken, until the 12th century, when it
appears, along with a multitude of other relics, in the capella imperatoris,
i.e. St. Mary of the Pharos 41 . That being so, I am a little surprised that
Ioli Kalavrezou should describe St. Stephen’s arm as «enriching imperial
ceremonial and strengthening the position of the emperor» 42 .
Furthermore, the very translation as described by Theophanes finds
no corroboration in other sources (for Proclus see below).This has been
shown at length by J. Wortley 43 , to whose demonstration we may add
one further text, namely the 9th-century 44 Brussels Chronicle, which
devotes particular attention to the arrival of saints’ relics at Constantinople.
For the reign of Theodosius II it records only that «the relics of the
36 THEODORUS ANAGNOSTES, Kirchengeschichte, ed. G.C. HANSEN, Berlin 1971 (Die
griechischen christlichen Schriftsteller der ersten Jahrhunderte, 54), p. 121, Nr.436.
Cf. CEDRENUS, I,p.619; ALEXANDER MONACHUS, Laudatio Barnabae, ed. P. VAN DEUN,
Turnhout-Leuven 1993 (Corpus Christianorum. Series Graeca, 26), p. 118 (in the
chapel of the palace).
37 Monumenta Germaniae Historica, Auct. ant., XI,p.191.
38 THEOPHANES, ed. DE BOOR, pp.299, 300, 769; CONSTANTINUS PORPHYROGE-
NITUS, De Cerimoniis, Bonn ed., I, pp. 196-197, 208, 212, 628.
39 See J. EBERSOLT, Le Grand Palais de Constantinople, Paris 1910, pp.52-53.
40 The contents of St. Stephen’s chapel, listed in De Cerimoniis,p.640, were limited
to «the great cross of St. Constantine», three sceptres and a few other objects of
secular ceremonial.
41 P. RIANT, Exuviae sacrae Constantinopolitanae, Geneva 1877, II, pp. 211-212,but
absent from the comprehensive listing by ANTONY OF NOVGOROD, Kniga Palomnik,
ed. Chr. LOPAREV, Pravoslavnyj Palestinskij Sbornik 51 (1899), pp. 18-19.
42 Helping Hands for the Empire, in Byzantine Court Culture from 829 to 1204, ed.
H. MAGUIRE, Washington, DC 1997, p.66.
43 The Trier Ivory Reconsidered, in Greek, Roman and Byzantine Studies 21 (1980),
pp. 381-394.
44 It is usually dated to the 11th century, but its final part is a bare list of emperors
from Basil I to Romanus III, no doubt added by a later copyist.The fact that the
Brussels Chronicle is the only Byzantine text to give the exact date of the first
Russian attack on Constantinople (860) strongly suggests that it was compiled soon
after that event.

A FAKE INSCRIPTION OF THE EMPRESS EUDOCIA 31
holy martyrs Stephen, Laurence and Agnes were deposited in the martyrion
of St. Laurence» 45 .The same information is repeated in a contaminated
manuscript of the Epitome of Theodorus Anagnostes, which adds
that the deposition took place on a 21 September, κα τελεται µέρι σήµερν
κεσε τ ατ µέρα µνήµη ατν 46 .The commemoration of 21 September
has not survived in our liturgical sources of the 9th-10th centuries
(Typicon of the Great Church, Synaxarion), which suggests that it
may have been of earlier date and had fallen into desuetude.
Theophanes, as we have seen, describes the despatch of St. Stephen’s
arm as a thanksoffering for the gold jewelled cross of Golgotha. This
piece of information has been taken at face value and connected with
the introduction of a new design on the reverse of gold coins, namely
that of a victory holding a long jewelled cross, which can be dated to
420 or 422 47 . Whether the long cross was meant to refer to that of Golgotha
I am unable to say, but there is a serious objection to this rapprochement:
the monumental jewelled cross erected on the rock of Golgotha is
prominently displayed in the famous apse mosaic of S. Pudenziana, dated
by art-historians to the 390’s 48 . Frolow, who was aware of the difficulty,
buried it in a footnote in which he boldly stated that the mosaic, or at
any rate its central part, cannot belong to the pontificate of Siricius (384-
399) or even to its restoration (?) in the early 5th century, but should be
ascribed to Hadrian I (772-795) 49 – a most unlikely view. We are faced
with a clear choice: either the mosaic is later than 420 or the story of
Theophanes is false.
An edifice of speculation based on the sermons of Proclus (bishop of
Cyzicus resident at Constantinople, 426-434, then Patriarch of Constantinople,
434-446) has been constructed by F. J. Leroy on the assumption
45 F. CUMONT, Anecdota Bruxellensia, I:Chroniques byzantines du MS 11376, Gand
1894, p.21.
46 PG 86/1, col. 216A. The passage in question has been excluded from
Hansen’s ed. He does not state his reasons for doing so. The text in PG goes back
to cod. Paris. gr. 1440, s. XVI.
47 This issue is dated by the legend VOT XX MVLT XXX. Strictly speaking,
the vicennalia of Theodosius should have fallen in 422, but his 15th anniversary was
celebrated in 415 and his 30th in 430. Discussion by Ph. GRIERSON - M. MAYS, Catalogue
of Late Roman Coins in the Dumbarton Oaks Collection, Washington, DC 1992,
pp. 142-143, with a preference for 420.
48 For the dating evidence see G. MATTHIAE, Mosaici medioevali delle chiese di
Roma, Rome 1967, pp.68-69.
49 Numismatique, cit., p. 86, n.4.

32
CYRIL MANGO
that Theophanes’ account was true 50 . He alleges that Proclus’ Homily 12
(on the Resurrection) was delivered in the chapel of St. Stephen built by
Pulcheria («en quelque sorte paroisse impériale») and contains a precise
reference to the mosaics that decorated it; that the Laudation of St.
Stephen (Homily 17) was spoken upon the translation of his relic to
Constantinople (thus excluding a date in 421) or its first anniversary; and
that two sermons wrongly attributed to John Chrysostom, namely PG
62, coll. 933-934 (on St. Stephen) and PG 50, coll. 715-720 (In proditionem
Iudae) are also by Proclus and belong to the same group. Of the four
texts in question we may eliminate Homily 17, which says nothing specific,
and In proditionem Iudae, which manifestly refers to Daphne, the
suburb of Antioch, not the Daphne palace at Constantinople 51 . Homily
12 and the short piece PG 63, coll. 933-934 do allude to the «virgin
Empress», i.e. Pulcheria, in vague terms.The former, which contains only
a fleeting reference to Stephen, lauds the Empress’s liberality (to new
baptizands?) and, incidentally, says nothing about mosaics 52 . The passage
that has been interpreted as pointing to St. Stephen’s palace chapel and
the relic kept in it occurs in PG 63, coll. 933-934 53 . Stephen, we are told
here, fought the Jews when he was alive; now that he is dead, he is
taking up arms against heretics (Nestorians?).The prize he received then
was similar to the one he is receiving now: «Stephen is in the palace, for
the Empress, who is a virgin, has stored him in her chamber [or in her
heart?]. Stephen is among magistrates, for he has made the Emperor’s
father the son of a man who has shown hospitality to strangers. Stephen
is in charity towards the poor, for he has become the bountiful and
faithful provider of widows and orphans» (ν ασιλείις Στέφανς θαλά-
µευσε γρ ατν ασιλς κα παρθένς παρ ρυσι Στέφανς τν γρ τ
ενδήσαντς παδα ασιλέως πεπίηκε πατέρα ν φιλπτωία Στέφανς
ηρν κα ρφανν πλσις γέγνε κα πιστς κνµς). Proclus (if it is
he) does appear to be alluding to certain contemporary cirumstances,
while representing Stephen as the champion of the poor, presumably for
50 L’homilétique de Proclus de Constantinople, Città del Vaticano 1967 (Studi e testi,
247), esp. p. 158.
51 It speaks of springs of water, groves of trees and the «Pythian demon» taking
fright at the Cross: PG 50, coll. 715-716.
52 Unless Leroy had in mind the sentence, PG 65, col. 788B, τν πίγειν ρανν
τν ρώµενν καλλώπισεν, whatever exactly that may mean.
53 HOLUM, Theodosian Empresses, cit., p. 104, n.116, uses this passage as his argument
massue to refute Wortley’s criticism (as in n. 43 supra).

A FAKE INSCRIPTION OF THE EMPRESS EUDOCIA 33
having been one of the seven chosen to minister to widows (Acts, 6.
1-5). For the emperor’s father the obvious candidate is the Persian
eunuch Antiochus, who acted as guardian of Theodosius II 54 , but we do
not happen to know whose son he was. The last sentence may refer to
the setting up of a charitable institution bearing St. Stephen’s name. But
what are we to make of the statement that Pulcheria «housed» (θαλάµευσε)
Stephen? Is it to be taken literally and, if so, does it necessarily
imply that she had deposited Stephen’s relic in the Imperial Palace rather
than in her church of St. Laurence built on her estate of Pulcherianae,
or should it be understood metaphorically? A parallel is provided by
Homily 12, PG 65, col. 788B: «She has mortified her own flesh against
the passions; she has housed (θαλάµευσεν) the crucified One in her
spirit». Here, certainly, the meaning is metaphorical.
To conclude this part of our discussion, I can see no firm grounds to
give credence to Theophanes’ story and several reasons for disbelieving
it. The only well-attested relic of St. Stephen in 5th-century Constantinople
is the one that was brought by Eudocia in 439. The conjunction
of Stephen and Laurence was not, of course, fortuitous.The two deaconmartyrs
were often thought as a pair and were represented side by side
in mural paintings. Their remains were allegedly laid in the same tomb
in the Pelagian basilica of S. Lorenzo fuori le mura, and there is a text of
uncertain date concerning the «coniunctio corporum... Stephani et Laurentii
facta Romae sub Theodosio iuniore» 55 . Seeing that Eudocia exercised,
so to speak, patronal rights over St. Stephen’s bones, one can readily
see why Pulcheria, who was not exactly on friendly terms with her,
might have wished to establish a prior claim on the Protomartyr – hence
the invention of the story related by Theophanes.
For some time Stephen’s relics remained at St. Laurence’s. In speaking
of them, Marcellinus Comes uses the present tense, «in basilica sancti
Laurentii positae venerantur». His Chronicle, as it survives today, extended
to 534 56 . There is also a neglected reference in the Miracles of Sts.
54 The chronology of Antiochus’ career is uncertain. According to J.R. MARTIN-
DALE, Prosopography of the Later Roman Empire, II, Cambridge 1980,p.102, he was dismissed
by Theodosius II in or shortly after 421. J. Bardill argues, however, that he
remained in service until 439:in G.GREATREX - J. BARDILL, Antiochus the Praepositus,
in Dumbarton Oaks Papers 50 (1996), p. 197.
55 Acta Sanctorum, Aug. II (1735), pp. 528-530.
56 It is, of course, possible that in this passage he neglected to update the first
edition of 518.

34
CYRIL MANGO
Cosmas and Damian (6th century?).The story concerns a Syrian woman
called Martha who, while a patient in the miraculous shrine, was subjected
to lesbian advances from one Christina, who was possessed by a
demon.The Saints appeared to Martha in a vision, bidding her not only
to avoid Christina, but to go home «for the sake of our brother St.
Stephen, for she happened to be the wife of a cleric of St. Laurence’s» 57 .
A synaxis of St. Stephen in St. Laurence’s on Wednesday of Easter week
is recorded in the 9th/10th-century Typicon of the Great Church 58 .By
that time, however, the cult of St. Stephen had migrated to his church at
Constanti(ni)anae, which became the venue of his synaxeis of 27 Dec.
(martyrdom), 2 Aug. (translation) and 15 or 16 Sept. (invention). Following
a suggestion by J. Pargoire 59 , this move has been attributed to the
initiative of the princess Anicia Juliana (d. 527/528). That has often been
stated as a fact, but is only a supposition, which may or may not be
right. I shall not discuss the matter here so as to spare the honorand a
great deal of tedium without the prospect of a clear solution.
CYRIL MANGO
57 L. DEUBNER, Kosmas und Damian, Leipzig-Berlin 1907, p.131.
58 Ed. J. MATEOS, Le Typicon de la Grande Église, Roma 1962 (Orientalia Christiana
Analecta, 165), II, p. 102.
59 À propos de Boradion, in Byzantinische Zeitschrift 12 (1903), pp. 488-490.

Pl. 1
Pl. 2